KR101322947B1 - Distributed caching of files in a network - Google Patents

Abstract

Distributed caching and download of files. A method is described that includes constructing a peer list that includes a list of potential peer servers among one or more computers connected in a network. The peer list contains only potential peer servers up to a predetermined number. Potential peer servers in the peer list are queried for a file or part of a file. A message is received from the peer server in the peer list indicating that the peer server has a downloadable file or part of the file. The computer system downloads the file or part of the file from its peer server.

Caching, file sharing, network, peer server

Description

Distributed caching of files on a network {DISTRIBUTED CACHING OF FILES IN A NETWORK}

Network computing allows a computer to request and receive software and file content from servers to be installed and / or stored on the computer. The file content may include, for example, a text file, a video file, a data file, an image file, and the like. The software may be requested and downloaded for various purposes, including system software updates, application software updates, and data file updates. For example, a security patch can be downloaded to update the operating system to prevent or remove an attack on a computer system. In addition, software can be downloaded to install new applications on the computer. In one example system, the software may be downloaded in the background so that the software may be downloaded while the computer and computer user perform other tasks. One example service with this feature is Background Intelligent Service (BITS).

Often, when software is published and available for download from a server, there are a number of computers in the network that want to acquire the software simultaneously or in a very short time. For example, if the software is a security patch, there is a desire to deliver the software to as many computers as possible in the shortest possible time to prevent malicious attacks on the computer or network. If the software is a new application for system wide distribution within a corporate campus, there may be a need to deploy the software system in a short time. If the software is a very anticipated release, many users may want to download the software to be one of the first users of the software. Large-scale deployment of software can require large amounts of server and network capacity. Such large downloads can be difficult, especially in a network environment where many branch offices are connected to the enterprise hub of the corporate headquarters. In this example, when the software is deployed from an enterprise hub, the network connection between the enterprise hub and the branch office may be updated at a particular time, even when the servers in the enterprise hub have enough functionality to provide software to many or all computers in the branch office. You can limit the number of computers you can.

There are several methods used to distribute the workload of large downloads. For example, in an enterprise environment, distributing new software can be accomplished by distributing the software only to a limited number of computers on the network at a given time. Alternatively, a caching proxy can be placed in the network to distribute software to a subset of computers in the network. Both of these solutions require careful planning for implementation. In addition, in the case of a caching proxy, additional computer hardware is added to the network, increasing the cost of the network.

While various issues have been identified in the background, the invention as claimed below does not necessarily cover any or all of the identified issues. This background is merely provided to provide a background for an exemplary environment in which the embodiments described herein may be practiced.

SUMMARY OF THE INVENTION [

One embodiment described herein includes a method for caching and sharing files. This method may be implemented, for example, in a computer system of a computer network including one or more computers connected by a network. The method includes building a list of peers that includes a list of potential peer servers among one or more networked computers. The peer list includes only potential peer servers up to a predetermined number. Potential peer servers in the peer list are queried for a file or part of a file. A message is received from one peer server in the peer list indicating that the peer server has a downloadable file or part of a file. The computer system downloads the file or part of the file from that peer server.

Another embodiment described herein includes a method for caching and sharing files. This method may be implemented, for example, in a computer system of a computer network including one or more computers connected by a network. The method includes sending a message that the computer system is available as a peer. The computer system receives a request to find out if the computer system has a file or part of a file. The computer system sends a message that the computer system has a file or part of a file. Sending a message that a computer system has a file or part of a file is performed based on dynamic opt-in / opt-out rules. The computer system receives a request to download a file. The computer system sends the file to the requesting system.

Another embodiment described herein includes a method for caching and sharing files. This method may be implemented, for example, in a computer system of a computer network including one or more computers connected by a network. The computer system sends a broadcast signal to the computer systems on the subnet to discover potential peers. The computer system receives a unicast, or alternatively multicast signal from one of the computer systems in the subnet. The computer system adds the one of the computer systems in the subnet to the peer list. The peer list includes only peers up to a predetermined number. The computer system sends a unicast signal to the one of the computer systems in the subnet to determine if the one of the computer systems has a file or portion of a file that can be downloaded. The computer system receives a unicast signal from the one of the computer systems indicating that one of the computer systems in the subnet has a file or portion of a file that can be downloaded. The computer system downloads a file or portion of a file from said one of the computer systems in the subnet.

This Summary of the Invention is provided to introduce a selection of concepts in a simplified form that are described in more detail in the Examples below. This Summary of the Invention is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.

In order to obtain the above-mentioned means and other features, more detailed description will be given with reference to the specific embodiment shown in the accompanying drawings. It is to be understood that the drawings only illustrate exemplary embodiments and therefore should not be considered as limiting the scope, which embodiments will be depicted and described with additional specificities and details through the use of the accompanying drawings.

1 is a block diagram of an example network.

2 shows a flow diagram illustrating a method for caching and sharing files from a client perspective.

3 shows a flow diagram illustrating a method for caching and sharing files from a peer server perspective.

One embodiment described herein includes an improved file distribution system that enables file distribution away from a centralized file server. In this example, file distribution and caching are moved to the edge of the network. This reduces the load on centralized file servers and helps distribute network traffic across the network, as opposed to concentrating network traffic on some centralized network connections. Specifically, in one embodiment computers use a peer to peer system to cache and share files. The computer system first builds a peer list that includes a list of potential peer servers among one or more computers connected to the network. In this embodiment, the peer list is limited to up to a predetermined number of peers. The system can then authenticate the peers in the peer list. The system can then query the peers in the peer list to find a peer that contains the file or portion of the file that the computer system needs. If a peer in the peer list responds that it has a file or part of a file available, it can download the file or part of the file from the peer that represents it. If no peer responds that it has a file or part of a file, the system requesting the file downloads the file from the central file server and indicates to other computer systems that it has the file or part of the file that it can download. Can be. In response to indicating that the file or portion of a file is available for download, other computer systems may add this computer system to a list of peers that includes a list of their potential peer servers, which is described in more detail below. It is explained. In particular, other peer systems may add the querying computer system to their peer list in response to a query to find a peer that includes a file or part of a file.

In one illustrative embodiment, a computer system may create a peer group from computers that are not interconnected by a central enterprise server or hub. To eliminate peer connections through expensive connections, such as cable modem connections, and the like, for example, a peer group may be formed at computers in a branch, common hub or router of an enterprise network. In this way, less network traffic and server resources are required to distribute the file to peer groups where the file can be cached and downloaded. In this way, the processing and network functions needed to distribute software and updates are moved from the centralized location of the network to the edge of the network, thus distributing computing and network requirements more evenly.

Referring now to FIG. 1, an exemplary topology 100 is shown. Topology 100 includes a number of computer systems 102, 104, 106, 108, 110. Computer systems 102-110 may be interconnected through any of a number of interconnecting means. For example, the computer system 102-110 may be an Ethernet network that conforms to IEEE 802.3, a wireless Ethernet connection such as those that conform to IEEE 802.11, a Bluetooth wireless network, a dial up networking connection, a cable modem connection, or the like. Can be interconnected through. In the illustrated embodiment, computer systems 102-110 are interconnected to form peer group 112. In this example, peer group 112 is a group of potential peer servers for a computer system, such as laptop computer system 106. In one embodiment, each computer system maintains a peer list that includes a list of potential peer servers from which the computer system can download files. The formation of peer groups will be discussed in more detail below.

1 also shows that peer group 112 is connected to central file server 114. Although the connection to the central file server 114 is shown as a single connection, in alternative embodiments, each of the computers 102-110 may have access to the central file server 114, or various connections may be used in other configurations. have. For example, a peer group may be formed at a branch in an enterprise enterprise, where the branch is connected to a central file server 114 in an enterprise hub. Various other interconnects may also be implemented.

In the example shown, the file server may store file 116 or a portion of a file. File 116 may be, for example, a software update, a security patch for an operating system, an application, a data file, or other software. File 116 may be downloaded or pushed to one of the computers, such as computer system 108. Computer system 108 may then act as a potential peer server with a cached copy of file 116. Other computer systems 102, 104, 106, and 110 can then request file 116 from potential peer server system 108. In this way, in some embodiments, network traffic on the connection 118 between the file server and the peer group 112 may be reduced. In addition, processing by the central file server 114 may be reduced.

As mentioned above, FIG. 1 shows that one of the computer systems is a laptop computer system 106 and that the peer group 112 is in the list of potential peer servers maintained by the laptop computer system 106. Other portable computers, such as laptop computers and hand-held devices, PDAs, cell phones, and the like, are portable according to their inherent nature and can therefore be located in physically and logically different parts of the network. The laptop computer 106 may, for example, be moved logically or physically to a location where the laptop computer 106 is not practical for downloading files from the peer group 112. For example, laptop computer 106 may be logically moved to a subnet other than peer group 112. Alternatively, laptop computer 106 may be physically moved to a location that requires excessive amounts of network traffic to allow laptop computer 106 to communicate with other members of peer group 112. In the embodiment shown in FIG. 1, laptop computer 106 creates another peer group, such as peer group 120 shown in FIG. 1. The laptop 106 maintains information about the first peer group 112 and the second peer group 120, such that the peer list of the first peer group 112 or the second peer group 120 or the laptop 106 is listed. The laptop 106 can receive the file when it is connected to any other peer group in.

When the computer system in peer group 112 receives the file 116, the computer system must confirm that the file is the correct file and is not corrupted. Determining that a file is an exact file and not corrupting may be done, for example, by comparing the file name, file size, hash value, and the like. For example, the computer system can receive information regarding the file name, file size, and hash value of the file. Hash values are values that are calculated using data in a file in order to generate a value in an extremely unlikely manner that any other data will produce the same hash value. The computer system can determine that the correct file is being received by comparing the file name, file size and / or hash value. The hash value can also be used to ensure that the file has not been corrupted or altered to malicious.

If the computer system detects that the correct file is not available, the file is corrupted, or the file has been changed to malicious, the computer system can request the file 116 directly from the central file server 114. Next, the computer system requesting the file 116 directly from the central file server 114 may advertise itself to the potential storage system containing the file 116.

In particular, in one embodiment, the peer group remains as described above and is specific to each computer system. Thus, peer group management is moved from a central location using the system without centralized control. This eliminates single point failure points for the entire network. In other words, the loss of one system with one peer list or one peer list does not prevent the entire system from distributing and caching files, and disabling the entire system is one central to managing peer groups. Occurs when a scheduled server is used.

Referring now to FIG. 2, a method 200 for caching and sharing files is shown. The method may be performed on a computer system of a computer network including one or more computers connected by a network.

2 illustrates the operation of building a peer list (step 202). The peer list includes a list of potential peer servers among a group of networked computers, such as the computer systems 102-110 shown in FIG. The peer list may include only potential peer servers up to a predetermined number. For example, the peer list may be designed such that only 100 or fewer potential peer servers are included in the peer list. In addition, computer systems 102-110 within a peer group need to be close in terms of networking. For example, computer systems 102-110 of peer group 112 may be required to be located at the same point in the enterprise network. In alternative embodiments, computer systems 102-110 of peer group 112 may be required to be located in a common hub, router, or switch. In some alternative embodiments, no peers may be required to be interconnected via expensive connections, such as connections between cable modems or high traffic connections. In addition, other logically close configurations may be implemented. For example, in one embodiment, a computer system may use a timing method to determine the nearness of computers. There are several ways that can be used alone or in combination to build a peer list (202). Examples of these methods will be described below.

One method 202 of building a peer list includes sending a solicitation request and receiving a response from the computers. In order to limit the number of potential peers responding, a communication induction request may be, for example, subnet broadcast or multicast to receive a response for a single subnet. However, even if only computers in a single subnet respond, the number of responses may still be excessive. Thus, to further limit the number of responses, the computer system sending the communication induction request may include in the communication induction request an indication of the number of peers required. Potential peers may be configured to respond using multicast after a random short interval and to hear multicast responses from other computers. If it is detected that the number of responses from other potential peers has already reached the indication of the required number of peers, the potential peer does not respond. In other words, if the required number of peers have already responded, the potential peer does not.

Several methods can be used to generate random short intervals. For example, in the simplest embodiment, a uniform random distribution may be used by each potential peer to determine how long to wait before responding to a communication induction request. An alternative method involves a non-linear distribution in which initially fewer potential peers respond and the longer the delay, the more potential peers respond. For example, one method of incorporating non-linear delay may include generating a random number between zero and one. The delay can then be defined by the equation

DELAY = MAX_DELAY * Log (C * X + 1) / Log (C + 1)

DELAY is the delay before responding, MAX_DELAY is the maximum amount of time given to receive the response, X is a random number, and C is any number such as 256. Other non-linear random delays can also be used.

In another example, the system can respond without delay and the computer system can add a peer to the peer list until there are enough peers to populate the peer list. In this example, the fastest responding peers may be those that are logically closest to the computer system sending the communication induction request. In this example, the computer system may take time to see how quickly potential peers respond. Only potential peers that respond the fastest are added to the peer list. Thus, a computer system can select peers that are most suitable for a peer group by accepting peers that can logically respond closest or fastest.

The request to direct communication may also include a group such as a trust group. Potential peers ignore requests associated with other trust groups. Even when a trust group is used for a request to induce communication, it may still be necessary to authenticate the peer as described in detail below. Trust groups are used as a way to sift responses from other trust groups in the same subnet so that duplicate communication induction requests in the subnet do not interfere with each other. In one embodiment, the communication derivation request may include an identification of the content required by the computer system. As described below, potential peers respond if they have the content of that identification and meet other opt-in constraints. This makes content retrieval more efficient because one content request reaches all potential hosts. This embodiment may be less private because unauthorized hosts may see which content is requested. In this embodiment, requesting a one-way hash of a URL, or a content hash, can be used to enhance privacy.

In another embodiment, building the peer list may include the computer system receiving a self-announcement with a multicast packet from a potential peer. Self-notification includes an indication that potential peers have a file or portions of files that can be provided to computers in a network. When such a notification is received, the computer system can add the potential peer to its peer list. Self-notification may be made regularly in one alternative embodiment. In one example, the notification may be made once per hour or multiple times per day.

Computer systems may self-notify when they have files to share, and when hardware characteristics, current processing load, or other characteristics allow the computer system to act as a peer server. Informing computer systems may receive file 116 from a central file server 114 or other peer in peer group 112. The computer system may have certain files that will not self-notify. This allows the computer system to protect certain files.

 In alternative embodiments, potential peers may announce when added to a network, such as when added to a network via multicast, for example. Next, if a peer is needed in the peer list, the computer can add this new computer to its peer list. Similarly, a computer system can add a peer to the peer list when receiving other types of multicast notifications and responses. For example, a computer system that listens for a response associated with the random time delay response embodiment described above may add to its peer list any computer systems that respond to inducing the original communication. Likewise, almost any response or notification can be used to determine which computers are available as potential peers in the peer list.

The computers in the peer list may or may not be verified. Validated computer systems are those that have been certified as outlined below. Peers in the peer list do not need to be validated immediately after being added to the peer list, but rather can be validated in close proximity to querying peers in the peer list for a file or part of a file (step 206), which is shown below. In more detail. However, once a peer in the peer list has been verified, the computer system and the peer share a key so that the verification does not need to be performed on subsequent file queries.

Part of the process of building a peer list may include removing an outdated item or updating the peer list to remove items that are no longer valid. For example, the computer system may remove items from the peer list after they have been in the peer list for a predetermined period of time. This ensures that old items have a limited lifespan. In another alternative embodiment, peer items may be removed in response to the multicast sent by the peer when the peer leaves the network. For example, when a peer prepares to leave the network, a "Bye" multicast one-way message is sent to other peers as best effort notification. If a "by" message is detected by the computer system, the peer sending the "by" message may be removed from the peer list. A "by" message may be less desirable because a malicious computer can use a "by" request to cause other computers to be removed from the peer list, thus allowing more computers to make requests to the central file server. to be. In another example, when an attempt is made to query a peer for a file, the peer may be removed from the peer list if an error is received indicating that the peer is no longer on the network. In one embodiment, if multiple queries to different peers of the peer list result in an error message, the computer system may discard all items of the entire peer list. This may be an indication that the portion of the network that the computer system is trying to access is no longer available. In another embodiment, if a peer is not verified, it may be removed from the peer list. This will be explained in more detail below, but in short, if the peer is not verified, this is an indication that the peer cannot be trusted. Similarly, if a file received from a peer does not pass certain security checks, such as hash comparison, the peer can be removed from the peer list. This will also be discussed in more detail below.

If a sufficient number of peers are removed from the peer list, the steps for building the peer list described above can be repeated. In particular, the steps described above can be used alone or together.

The computer system may maintain one or more peer lists. For example, in FIG. 1, an example of a laptop computer system 106 is shown. As described, laptop computer system 106 may be physically and / or logically relocated. In one embodiment, laptop computer 106 may be logically relocated and exist in a different subnet than first peer group 112. The laptop computer system 106 may detect that it has moved to a new subnet, and in response to detecting that the laptop computer system has moved to a new subnet, the laptop computer system 106 may construct a new peer list for the new subnet. To do this, the above-described steps for building a peer list can be performed. If the laptop computer system 106 is regularly a member of certain subnets, the laptop computer system 106 maintains a peer list for these subnets to avoid having to perform the steps of building a peer list each time it associates with the subnets. Can be. The number of different peer lists maintained by the computer system may be a statically defined number, or alternatively the number may be a configurable number by a user or a network administrator.

In particular, there may be a situation where a computer system moves to a new subnet but does not form a new peer group. For example, if a laptop computer moves to an internet cafe, the laptop computer probably does not form a new peer group. Communication with available peers can be too risky in this and other situations.

The computer system can also be configured to build a peer list when first added to the network. This can be done by any new computer system, but is not limited to laptop systems.

Referring again to FIG. 2, an operation (step 204) of authenticating a computer in the peer list is shown. Authentication involves verifying that the peer is a trusted machine. In one embodiment, the computer system may verify that the peer is a member of a particular security group. Validation may, for example, use a protocol to allow both computer systems and peers to perform mutual-authentication. In this way, by verifying the identity of both the computer system and peer, it prevents "spoofing" each other to believe that peers in the computer system or peer list are in the same security group. One protocol for this mutual authentication may be Kerberos. If the peer fails verification, the peer can be removed from the peer list.

Once the peer has been verified, the computer system and the peer share the key so that subsequent verifications do not have to be performed before subsequent queries. The key may have an expiration period to make trust verification effective. This expiration period can be, for example, one day.

After the peer is verified, FIG. 2 illustrates an operation (step 204) of querying peers in the peer list for a file or portion of a file. Querying a peer in the peer list may, in one example embodiment, send a unicast message to each peer in the peer list until one of the peers responds that it has the file or part of the file it is looking for. It includes sending. As used here, note that unicast is a network query to a specific computer. Unicast described here does not necessarily require the use of a datagram protocol, but may use other protocols such as TCP and other protocols. In an alternative embodiment, the computer system may transmit a set of overlapping unicast messages. For example, the computer system may send a unicast message to peers in the peer list before receiving a response from peers that previously sent the unicast message.

The message sent in the querying step may include various parameters for identifying the file needed by the computer system. For example, the message may include one or more of a file name, file size, file creation date, and / or hash of the file. Specifically, by using a hash of the file, the system can detect that the correct file is available. Using the hash, the computer system can detect whether the available file is an updated file, a duplicate name of another file, and the like.

After the computer system queries the peers of the peer list for a file or portion of a file (step 206), the method shown in FIG. 2 may inform the peer list that the peer server has a downloadable file or portion of a file. The operation of receiving from is shown (step 208).

When deciding whether a peer will act as a peer server, the peer may refer to various policy rules. Additionally, even when a peer has an available file, that peer may choose not to act as a peer server according to policy rules. Such policy rules may take into account the processing power of a computer system. For example, a powerful desktop or server class computer may include policy rules that allow it to act as a peer server for more peers, eg, a laptop computer such as laptop computer system 106. Other rules may relate to the amount of available machine resources. For example, policy rules may limit the ability to act as a peer server of a computer system when more than 10% of the computer system resources are used. System resources may be a measure of the greater of the system's CPU load and disk input / output load. The laptop computer may also refer to a policy that adjusts the number of peers for which the laptop computer system 106 acts as a peer server, based on the operating state of the laptop computer system 106. For example, if laptop computer system 106 is connected with a docking station, the laptop computer may be configured to face more peers than when not connected with the docking station. Furthermore, when laptop computer system 106 is connected via a wired connection, laptop computer system 106 may encounter more peers than when connected with peer computers in a wireless manner. Also in another example, the number of peers allowed to connect with laptop computer system 106 may be defined by a policy related to battery level. For example, when the laptop is running on battery and the power is less than 50%, the laptop computer system 106 may follow a policy of disallowing further peer computer connections.

If the peer decides to act as a peer server, the peer server sends a message back to the computer system, which is received by the computer system (step 208).

After the computer system receives a message from a potential peer server in the peer list that indicates that the potential peer server has a downloadable file or portion of a file, the computer system downloads (step 210) the file or portion of the file. The download can be in any suitable manner. In particular, as implied herein, a computer system may download a portion of a file rather than an entire file. Different parts of the file may be obtained from other peer servers. In some circumstances, this may increase the speed at which large files are downloaded.

In one embodiment, the method 200 further includes verifying the file or portion of the file (step 212). Validation confirms that the file is the correct file, the file has not been changed to malicious and / or the file is not corrupted. This can be done, for example, by calculating a hash of a file or part of a file. If the calculated hash value does not match the expected hash value, the computer system discards the file or part of the file. In addition, the computer system may place the peer server that provided the file in a quarantine list indicating that the peer server is a suspect. In alternative embodiments, the peer server may be removed from the entire peer list. The computer system can then attempt to get the file from another peer in the peer list or from a central file server. The computer system can also inform the peer server that provided the file that there is a problem with the file. This allows the peer server to discard the file. Other computer systems may also be informed about the suspected nature of the file.

Referring now to FIG. 3, an embodiment is shown that illustrates a method 300 for caching and sharing files in terms of potential peer servers. The method shown in FIG. 3 may be implemented in a computer system of a network, where the network includes one or more computers connected by a network.

The method 300 includes transmitting a message that a computer system is available as a peer (step 302). Sending a message that the computer system is available as a peer (step 302) may be performed using multicast on the subnet when the computer system has a file to share. In an alternate embodiment, sending a message that the computer system is available as a peer may be performed using multicast or unicast in response to receiving a request to direct communication from a system seeking a peer computer. . When sending a message that the computer system is available as a peer (step 302) is performed in response to receiving a communication induction request, the computer system uses the procedure described above in conjunction with the technique of FIG. random delay) to respond.

The method 300 also includes receiving a request to discover if the computer system has a file or a portion of a file (step 304). The received message may be a unicast request from a system that includes a computer system in its peer list. In alternative embodiments, the request may be a duplicate unicast as described above. Multicast requests can also be used when no security is required for the file.

The method 300 also includes transmitting a message that the computer system has a file or a portion of a file (step 304). Sending a message that the computer system has a file or part of a file (step 304) is performed based on dynamic opt-in / opt-out rules. For example, in one embodiment, policy rules may take into account the processing power of a computer system. Powerful desktop or server class computers can have policy rules that allow them to act as peers to more peers than other less powerful systems. Other rules may relate to the amount of available machine resources. For example, policy rules may limit the ability of a computer system to act as a peer server when more than 10% of computer system resources are used. System resources may be a measure of the greater of the system's CPU load and disk input / output load. In an alternative embodiment associated with a laptop computer, the laptop computer may refer to a policy that adjusts the number of peers for which the laptop computer system 106 acts as a peer server based on the operating state of the laptop computer system 106. have. For example, if laptop computer system 106 is connected with a docking station, the laptop computer may be configured to face more peers than when not connected with the docking station. Furthermore, when laptop computer system 106 is connected via a wired connection, laptop computer system 106 may encounter more peers than when connected with peer computers in a wireless manner. Also in another example, the number of peers allowed to connect with laptop computer system 106 may be defined by a policy related to battery level. For example, when the laptop is running on battery and the power is less than 50%, the laptop computer system 106 may follow a policy of disallowing further peer computer connections.

The method 300 also includes receiving a request to download a file (step 308). Further, the method 300 includes transmitting a file to the requesting system (step 310). This may be performed by a computer system acting as a peer server to other systems in the network.

Referring now to FIG. 4, an embodiment is shown that illustrates a method 400 for caching and sharing files, including various messaging formats. The method shown in FIG. 4 may be implemented in a computer system of a network including one or more computers connected by a network.

The method includes the computer system sending a broadcast signal to computer systems in the subnet to find a potential peer (step 402). The broadcast signal may indicate a request for a predetermined number of responses. In alternative embodiments, the broadcast may include various parameters to allow the computer system to calculate any response to the broadcast signal. For example, the parameters may allow for timing of the response, evaluation of the hardware characteristics of the responding system, and the like.

The computer system then receives a signal from one of the computers in the subnet (step 404). The signal may include a list of files that may be provided, characteristics of the system transmitting the signal, and the like. The signal may be a multicast signal, or in some cases a unicast signal.

The computer system adds the one of the computers in the subnet to the peer list (step 406). The peer list includes only peers up to a predetermined number. For example, a peer list may contain only 100 peers or less. This allows the computer system to limit the amount of processing and network usage required to find a file or portion of a file for download.

The computer system sends a unicast signal to one of the computer systems in the subnet to discover if one of the computer systems has a downloadable file or a portion of the file (step 408). The computer system can repeat this process for other computer systems to find where the file or part of the file is located. In particular, unicast signals may overlap, resulting in simultaneous unicast from the same computer system.

The computer system receives a unicast signal from one of the computer systems in the subnet, indicating that one of the computer systems in the subnet has a downloadable file or portion of a file (step 410).

The computer system can then download the file or portion of the file from one of the computer systems in the subnet (step 412).

Those skilled in the art will appreciate that the present invention provides a network computing environment through various types of computer system configurations, including personal computers, hand-held devices, multiprocessor systems, microprocessor-based or programmable consumer electronics, network PCs, mini computers, mainframe computers, and the like. It will be appreciated that it can be run from. The invention may also be practiced in distributed computing environments where tasks are performed by local and remote processing devices that are linked through a communications network (by hardwired link, wireless connection, or a combination of wiring and wireless connections). have. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

Embodiments within the scope of the present invention also include a computer readable medium for carrying or storing computer executable instructions. Computer readable media can be any available media that can be accessed by a general purpose or special purpose computer. By way of example, such computer-readable media may comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage device, or necessary program code means in the form of computer executable instructions or data structures. Including, but not limited to, any other media that may be used to store or store, and accessible by a general purpose or special purpose computer. When information is transmitted to a computer via a network or other communication connection (wired, wireless, or a combination of wires or wireless), the computer properly views this connection as a computer readable medium. Thus, all such connections are suitably referred to as computer readable media. Combinations of the above should also be included within the scope of computer-readable media. Computer-executable instructions include, for example, instructions and data that cause a general purpose computer, special purpose computer, or special purpose processing device to perform a particular function or group of functions.

Although the invention has been described in specific language in terms of structural features and / or methodological acts, it is to be understood that the invention defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms for implementing the claims.

The invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. The described embodiments are to be considered in all respects only as illustrative and not restrictive. Accordingly, the scope of the invention is indicated by the appended claims rather than the foregoing description. All changes that come within the meaning and range of equivalency of the claims are to be embraced within their scope.

Claims (20)

A method for caching and sharing files in a computer system in a computer network including one or more computers connected over a network, the method comprising: Constructing a peer list comprising a list of potential peer servers among the one or more computers connected to the network, the peer list comprising a potential peer server of less than a predetermined number; Querying a potential peer server in the peer list for a file or part of a file, Receiving a message from a potential peer server in the peer list that the potential peer server has a downloadable file or a portion of a file; Downloading the file or portion of a file from the potential peer server, The step of building the peer list, (a) sending a solicitation request to one or more computers including an indication of the number of potential peer servers required; (b) receiving a message response from the computer, the computer waiting for a random interval of time listening to a multicast response from another computer, and the computer waiting from the other computer. Detects a response to determine the number of responses from the other computer, the computer sends a response only if the computer does not detect a number of responses corresponding to the required number of potential peer servers, and the computer Does not send a response from the other computer if it detects the number of responses indicated in the number of potential peer servers required -and, (c) based on the message response received from the computer, adding, as the potential peer server to the peer list, the computer that sent the response up to the predetermined number, wherein the potential peer server is the potential for the computer system. Added to the peer list according to the logical closeness of the peer server; (d) removing a potential peer server from the peer list; (e) repeating steps (a) to (c) if a predetermined number of potential peer servers have been removed from the peer list. File caching and sharing methods. The method of claim 1, The computer waiting for a random time interval is a logarithmic random, calculated as DELAY = MAX_DELAY * Log (c * X + 1) / Log (C + 1) to determine the amount of time to wait before a response. delay), where DELAY is the delay before the response, MAX_DELAY is the maximum amount of time given to receive the response, X is a random number, and C is any number File caching and sharing methods. The method of claim 1, The transmitting and receiving steps consist of a computer on the same first subnet. File caching and sharing methods. The method of claim 3, Releasing a connection with the first subnet, establishing a connection with a second subnet, and sending a request request to establish a second peer list; Receiving a response from a computer on the second subnet File caching and sharing method further comprising. The method of claim 1, The step of building the peer list, Receiving a self-announcement with a multicast packet from the computer, the self-announcement comprising an indication that the computer has a file that can be provided to other computers in the network; and Adding the computer to the peer list File caching and sharing method that includes. The method of claim 1, The request for request includes the identification of a trust group. File caching and sharing methods. The method of claim 1, Determining that the file or part of the file is damaged and downloading the file or part of the file from a central server in response to determining that the file or part of the file is damaged. File caching and sharing methods. The method of claim 7, wherein Determining that the file or part of the file is corrupt includes referencing a hash value File caching and sharing methods. The method of claim 1, In response to determining that the file or part of the file cannot be provided from any computer in the peer list, and in response to determining that the file or part of the file cannot be provided from any computer in the peer list, Downloading the file or portion of a file from a central server further; File caching and sharing methods. The method of claim 1, Authenticating a computer in the peer list before querying a computer in the peer list for a file or portion of a file; File caching and sharing methods. A method for caching and sharing files in a first computer system in a computer network including one or more computers connected over a network, the method comprising: Receiving a solicitation request, the solicitation request including an indication of the number of potential peers being sent and requested to one or more potential peers; and Waiting for a random time interval while listening to multicast responses from other potential peers, detecting responses from the other potential peers to determine the number of responses from the other potential peers, and corresponding to the number of potential peers required Transmitting a response only to the request request if the number of responses is not detected, and not transmitting a response from the other potential peers as indicated in the number of potential peers required; After the random time interval has passed, the number of responses corresponding to the required number of potential peers has not been detected and, as a result, the message indicating that the first computer system is available as a potential peer is received. 1 transmitting to one or more computer systems on the same subnet as the computer system, As a result of sending a message to the one or more computer systems on the same subnet as the first computer system that the first computer system is available as a potential peer, whether the first computer system has a file or part of a file Receiving a request from a requesting system for making a bet, In response to the request to find out whether the first computer system has a file or part of a file, sending a message to the requesting system that the first computer system has a file or part of a file; Receiving a request from the requesting system to download the file as a result of sending a message that the first computer system has a file or part of a file; In response to receiving the request to download the file, transmitting the file to the requesting system. File caching and sharing methods. A method for caching and sharing files in a computer system in a computer network including one or more computers connected over a network, the method comprising: (a) transmitting a broadcast signal to a computer system on the subnet to find a potential peer, the broadcast signal including an indication of the number of potential peers required; (b) receiving a response signal from one of the computer systems in the subnet, wherein one of the computer systems waits a random time interval while listening to a multicast response from another computer system, and one of the computer systems Detects responses from computer systems to determine the number of responses from other computer systems, and one of the computer systems only detects the number of responses from other computer systems corresponding to the required number of potential peers. Transmit the response signal and do not transmit the response signal if it detects the number of responses indicated in the required number of potential peers from another computer system; (c) optionally adding the one of the computer systems of the subnet to a peer list based on a logical proximity of one of the computer systems, the peer list comprising up to a predetermined number of potential peers; and (d) removing potential peers from the peer list; (e) optionally adding a computer system of the subnet to the peer list when a predetermined number of potential peers have been removed from the peer list; Transmitting a unicast signal to the one of the computer systems of the subnet to determine if the one of the computer systems has a downloadable file or a portion of the file; Receiving from the one of the computer systems of the subnet a unicast signal indicating that the one of the computer systems of the subnet has the downloadable file or a portion of the file; Downloading the file or portion of a file from the one of the computer systems in the subnet File caching and sharing method that includes. delete delete delete delete delete delete delete delete

Images